乳腺肿瘤相关文献

CancerTreatmentReviewsjournalhomepage:www.elsevierhealth.com/journals/ctrvTUMOURREVIEW

Intravesicaltherapyforsuperficialbladdercancer:Asystematicreviewofrandomisedtrialsandmeta-analyses

MikeD.Shelleya,*,MalcolmD.Masonb,HowardKynastoncaCochraneUnit,ResearchDepartment,VelindreNHSTrust,Whitchurch,Cardiff,WalesCF142TL,UKDepartmentofOncologyandPalliativeMedicine,UniversityofWales,Cardiff,UKcDepartmentofUrology,UniversityHospitalofWales,Cardiff,UKbarticleinfosummary

Background:In2002therewereestimatedtobe357,000newcasesofbladdercancerworldwideand145,000deathsmakingbladdercancerthe9thmostcommonmalignancyglobally.Atdiagnosis,60–80%oftumoursaresuperficialandendoscopicresectionistheinitialtreatmentforthisdisease.Inpatientswithlow,mediumorhighriskdisease,about20%,40%and90%,respectively,willdeveloptumourrecurrence.Todelayorpreventrecurrence,intravesicaltherapyisroutinelyused.CommonlyusedintravesicalagentsincludeimmunotherapywithBCGandchemotherapywithcytotoxicssuchasMitomycinC,Adriamycin,EpirubicinandGemcitabine.However,controversyexistsastowhichagentandscheduleshouldbeused.Methods:Anoverarchingsearchoftheliteraturewasusedtoidentifyrelevantstudiestoassesstheclin-icalbenefitofintravesicaltherapyandprovideclinicalguidanceinacomprehensivesystematicreviewofrandomisedtrialsandmeta-analysesofintravesicaltherapyforsuperficialbladdercancer.Findingsandinterpretationthesearchidentifiedover80randomisedtrialsand11meta-analyses.Theextensiveevi-dencesuggeststhatanimmediatepost-operativeinstillationofachemotherapeuticagent,suchasMito-mycinCorEpirubicin,iseffectiveinreducingtumourrecurrence.Inintermediateorhighriskpatients,furtherintravesicalinductionandmaintenancetherapywithBCGisrecommended.Methods:IntravesicalchemotherapywitheitherMitomycinCorEpirubicinwouldbeanoptionforthosepatientsfailingorwhoareunsuitableforBCGtherapy.IntravesicalBCGissuperiortochemother-apyintermsofcompleteresponseanddisease-freesurvival.However,thereisnoconclusiveevidencethatoneagentissuperiorintermsofoverallsurvival.Ó2009ElsevierLtd.Allrightsreserved.Articlehistory:Received6October2009Receivedinrevisedform1December2009Accepted5December2009Keywords:SuperficialbladdercancerIntravesicaltherapySystematicreviewRandomisedtrialsMeta-analysesIntroductionIn2002therewereestimatedtobe357,000newcasesofblad-dercancerworldwideand145,000deaths.1Thismakesbladdercancerthe9thmostcommonmalignancyandthe13thmostcom-moncauseofcancerdeathglobally.Overthelastdecadetheinci-dencehasbeenincreasingalthoughthemortalityrateappearstostabilisingorfalling.Itisdiagnosedmorefrequentlyinmenthanwomen(ratio4:1)whichmayreflecttheexposuretoriskfactorssuchasindustrialdyesandsmoking.Atdiagnosis,60–80%oftumoursaresuperficialandconfinedtotheurotheliumand/orthelaminapropria.Thesecomprisepapil-larytumours,designatedTa(urothelial)andT1(extendingintothelaminapropria)orcarcinomainsitu(CIS),aflaterythmatouslesionwithapropensitytoprogresstomuscle-invasivedisease.BothpapillarytumoursandCIScommonlypresentconcurrently.*Correspondingauthor.Tel.:+44(0)2920316909;fax:+44(0)2920529625.E-mailaddress:mike.shelley@velindre-tr.wales.nhs.uk(M.D.Shelley).0305-7372/$-seefrontmatterÓ2009ElsevierLtd.Allrightsreserved.doi:10.1016/j.ctrv.2009.12.005Symptomsofearlybladdercancerthatalertpatientstoseekmedicaladviceincludehaematuriaandurinaryfrequency.Thosesuspectedofhavingbladdercancerundergocystoscopyandtrans-urethralresection(TUR)fordiagnosis,and,ifpossible,totalendo-scopicresectionofvisibletumour.Forlowgradepapillary(pTaG1)tumoursthismaybetheonlytreatmentrequired.However,tu-mourrecurrenceisamajorprobleminpatientswithhighergradeTaandT1lesions.AtoneyearpostTURabout20%ofpatientswithlowriskdiseaseand40%ofthosewithmediumriskdiseasewilldeveloptumourrecurrence.Patientswithhighriskdiseasewillex-pressanevengreaterrecurrencerate(90%)at1–2yearsfollowingTUR.Thetherapeuticapproachmostcommonlyusedisintravesicaltherapy,wherebyagentsareinstilledintothebladdertodelayorpreventtumourrecurrence.Intravesicalagentsthatreducetumourrecurrencemaypossiblyreducethedevelopmentofdiseaseprogression.Thereareanumberofimmunologicalagentsandseveralche-motherapeuticsthathavebeenadministeredintravesicallyforthemanagementofsuperficialbladdercancer.Immunotherapy196M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205withintravesicalBacillusCalmette–Guerin(BCG),anattenuatedformusedasavaccineagainsttuberculosis,wasfirstreported2onin1976andsincethenmanyreportshaveconfirmedits’effi-cacyintheprophylaxisoftumourrecurrence3anddiseaseprogres-sion.4Intravesicalchemotherapyhasbeenusedforanequallylongtimeandalthoughpublisheddataindicateaneffectontumourrecurrence,5,6aclearbenefitintermsofdiseaseprogressionhasyettobeestablished.Herewepresentasystematicreviewofpub-lishedrandomisedtrialsandmeta-analysesreportingontheeffi-cacyandside-effectsofintravesicaltherapyinthemanagementofsuperficialbladdercancer.SearchstrategyMEDLINEwassearchedfrom1966–2009forrandomisedtrialsofintravesicaltherapyforsuperficialbladdercancerusingboththesaurusandtextwordtermsfor‘bladderneoplasm’or‘intraves-ical’and‘immunotherapy’,‘chemotherapy’and‘meta-analysis’andtranslatedfortheCochraneLibrary,EMBASE,SCOPUS,ASCO,andBIOSIS.TheCochranesearchfilterwasusedtoidentifyrandomisedtrials.Therewerenolanguageorlocationrestrictions.Abstractsofpaperswerereadandifconsideredrelevanttotheaimsofthisre-view,thefullpaperswereobtainedandreviewed.Thebibliogra-phyofeacharticlewasusedtoidentifyfurtherrelevantrandomisedstudies.IntravesicalBCGIntravesicalBCGisaneffectivetherapeuticoptionforpatientswithsuperficialbladdercancer.AnumberofrandomisedtrialshaveevaluatedthebenefitofadjuvantintravesicalBCGfollowingTURcomparedwithTURaloneinintermediateandhighriskpa-tients.ThestrainsofBCGincludedTokyo,ConnaughtandPasteur,withdosesrangingfrom75mgto150mganddwelltimesof1–2h.Asystematicreviewandmeta-analysis7(Table1)ofthesedataindicatedthatBCGadjuvanttoTURsignificantlyreducedtheriskofrecurrenceat12monthsby67%comparedtoTURalone.BCGwasassociatedwithanumberofcomplicationsincludingurinaryfre-quency(71%),cystitis(67%),fever(25%)andhaematuria(23%).Aseparatemeta-analysis8combiningbothrandomisedtrialsandret-rospectivestudies,reportedtumourrecurrencewassignificantlylesswithBCGcomparedtoTURalone(oddsratio0.35%,95%con-fidenceinterval0.20–0.59;p<0.001).Anadditionalmeta-analysis4suggeststhatBCGsignificantlyreducestheriskofdiseaseprogres-Table1

Meta-analysesofintravesicaltherapyforsuperficialbladdercancer.Study/aimsShelleyameta-analysisofpublishedRCTstocompareincidenceoftumourrecurrencefollowingTURalonewithTURplusintravesicalBCGHan8meta-analysisofpublisheddataonBCGandtumourrecurrencerateSylvester4determineeffectofintravesicalBCGonriskofdiseaseprogression.Pan17determineroleofmaintenanceBCGtherapyinT1G3tumoursBohle3tocomparerecurrenceandtoxicityofintravesicalBCGwithMMCinTa/T1bladdercancer.Shelley30comparetheefficacyofintravesicalBCGwithMMCBohle33tocompareriskofprogressionofintravesicalBCGwithMMCinTa/T1bladdercancerMalmstrom9tocomparetheefficacyofBCGwithMMCintermsofrecurrence,progressionandsurvivalHuncharek5determineimpactofintravesicalchemotherapyontumourrecurrencefollowingcompleteTUR7Includedstudies585Patientsfrom6RCTs(281TURalone,304TURplusBCG).4differentstrainsofBCGincluded,withdosesof78–180mginstilledover1–2h25Trials(RCTsandretrospectivestudies)comparingBCG(n=2,342)versusTURorothernon-BCGintravesicaltherapies(n=2425)24RCTscomparingTUR+BCGversusTUR+non-BCGtreatment(4863Ta/T1/CIS)13RCTSorcontrolledtrialscomparingmaintenanceBCG(n=915)withnomaintenance(n=733)andreportingrecurrencedata11Controlledtrials(n=2,799)recruitingintermediate/highriskpatientsreceivingBCGorMMCResultsTumourrecurrencewassignificantlylessinpatientsreceivingTURplusBCG(OR0.33,95%CI0.21–0.43,p<0.0001)RecurrencewassignificantlylessinBCGgroup(40.5%versus49.7%,p<0.0001)BCGsignificantlyreducedriskofprogression(rate:BCG9.8%,non-BCG13.8%)OR0.73(95%CI0.60–0.,p=0.001)41%maintenancegrouprecurredcomparedto45%incontrolgroup(oddsratio0.58,95%CI0.41–0.83,p=0.003)39%ofBCGgroupand46%inMMCgrouprecurred(OR0.56,95%CI0.38–0.84,p=0.005).MaintenancetherapyappearedtobeimportantinBCG’ssuperiorityInhighriskpatientsBCGsignificantlyreducedriskofrecurrence(31%reductioninprobabilityofrecurrenceperunittime,p<0.001).Nodifferenceinprogressionrate(BCG7.7%,MMC9.4%).However,BCGsuperiorwhenBCGmaintenancegroupcomparedtoMMC(OR0.66,95%CI0.47–0.94,p=0.02)BCGplusmaintenancesuperiortoMMCforrecurrence(p<0.001)butnosignificantdifferencewasobservedinratesofprogressionandsurvivalSignificantreductioninrecurrencewithintravesicalchemotherapy.Sub-analysisindicatedimprovedeffectwithlongerschedulesAmeta-analysisofpublisheddatafrom1527Ta/T1patientsfrom6RCTs(834BCGversus693MMC)9Controlledtrials(7prospective,1retrospective,1observational)comparingBCGwithMMC(2410patients).AnIPDmeta-analysisfrom9RCTs(2820patients)comparingBCGwithMMC.11RCTs(3730patientsTa/T1G1–G3tumours).ComparedTURversusTUR+intravesicalchemotherapy(ADR,MMC,EP,thiotepa,peplomycin,neocarbarzine,mitoxantrone).Treatmentvariedfromasingleinstillationtoa2yearschedule4EORTCand2MRC(2,535patients)prophylacticRCTsinprimaryorrecurrentTa/T1patientsassessingTURwith(1629)orwithout(906)intravesicalchemotherapy(thiotepa,VM-26,ADR,epodyl,Epirubicin,MMC,pyridoxine)9RCTs700patientswithCIS.ComparedBCGwithintravesicalchemotherapy(MMC,EP,ADR)Pawiniski6toevaluatetheimpactofprophylacticchemotherapyagentsfollowingTUR,onrecurrence,progressionandsurvival.Anindividual-patient-datameta-analysisSylvester35assesstheefficacyoflong-termorshort-termBCGandchemotherapyforCISSylvester73assesstheeffectofasingleimmediateintravesicalinstillationonriskofrecurrenceAdjuvantchemotherapysignificantlyreducedtherisktherecurrenceandincreasedthedisease-freeintervalP<0.01).Therewasnobenefitfordiseaseprogressionorsurvival68%CompleteresponseonBCGcomparedto51%onchemotherapy(p=0.0002).47%onBCGdisease-free,26%onchemotherapy(p<0.0001)BCGsuperiorforCISSingleinstillationsignificantlyreducedriskofrecurrence(OR0.61,95%CI0.49–0.75,p=0.0001).Moreeffectiveforsingletumours7RCTscomparingTURaloneversusTUR+singlepost-operativecytotoxicinstillation(MMC,EP,thiotepa,pirarubicin,1476Ta/T1).TUR–transurethralresection,RCT–randomisedcontrolledtrial,BCG–BacillusCalmette–Guerin,OR–oddsratio,CI–confidenceinterval,CIS–carcinomainsitu,MMC–MitomycinC,EP–Epirubicin,ADR–Adriamycin,CR–completeresponse,IPD–individual-patient-data.M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205197sioncomparedtoTURplusnon-BCGtherapy.ItisclearfromthesemanystudiesthatadjuvantBCGissignificantlymoreactivethatsurgeryalone.BCGscheduleTooptimisetheeffectivenessofintravesicalBCG,randomisedstudieshaveinvestigatedvaryingthedose,durationandconcom-itantnon-intravesicaladministrationofBCG.ThestandarddoseofintravesicalBCGcommonlyusedatpres-entis81mgdryweightor120mgwetweight,althoughforaccu-racyandcomparativeanalysis,thedoseshouldalsobeexpressedincolonyformingunits.Onlyafewrandomisedstudieshaveinves-tigatedthedoseeffectofintravesicalBCG.TheSpanishCUETOgroupofresearchers(ClubUrologicoEspanoldeTratamientoOncologico)reporteddatacomparingstandardandlowdosesofintravesicalBCGinTaG2-3/T1G1-3tumourswithorwithoutcon-comitantCIS.9FivehundredpatientswererandomisedfollowingTURtoeither81mgor27mgBCGgivenintravesicallyfor6weeksthenanother6instillationseverytwoweeks.Twenty-ninepercentofpatientsrecurredinthestandarddosearmcomparedto31%inthelowdosearmwithnosignificantdifference(p=0.586).Pa-tientsconsideredhighriskforrecurrencefaredbetteronthestan-darddose,althoughthedifferencewasnotsignificant(29.5%versus36.7%recurrences,p=0.14).Asimilarnumberofpatientsineachdosegroupdevelopeddiseaseprogressionandnosignifi-cantdifferenceinoverallsurvivalwasobservedat5years,althoughthestandarddosewasmoreeffectiveinmutlifocaldis-ease(p=0.048).Thenumberofpatientsexperiencingnotoxicitywasgreaterinthelowdosearmbutthosewithseveretoxicitywereevenlymatchedbetweengroups.Thesedatasuggestthatthestandarddoseshouldbeusedforhighriskpatientsandthelowdoseforintermediateriskdisease.Recruitingonlyhighriskpatients(T1G3andCIS)confirmedtheequivalenceofthetwodosesintermsofrecurrenceandprogression10suggestingthatthestandarddoseshouldonlybeusedforcasesfailingthe27mgdose.LoweringthedoseofBCGto13.5mgresultedinsignificantlylessefficacy.11Itisstillunclearwhattheoptimaldoseis,andcon-sideringthatBCGtherapyisactivebutinducessignificantside-ef-fects,moreresearchneedstobedoneinthisarea.However,arecentlyclosedEORTCtrial30962isaddressingthisissue.Thisran-domisedstudycomparedstandarddoseBCGwithlong-termmain-tenanceversusBCG1/3dosewithlong-termmaintenanceversusBCGstandarddosewithshort-termmaintenanceinintermediateandhighriskpatientswithTa–T1papillarycarcinomaoftheblad-der.TheresultsofthistrialareeagerlyawaitedandwillprovideusefuldataforoptimisingthedoseofintravesicalBCG.ThebenefitofmaintenancetherapyfollowinginitialBCGinduc-tionhasbeenthefocusofanumberofrandomisedstudies.12–16InthesestudiespatientswererandomisedtoreceivenofurtherBCGtherapyfollowinginductionormaintenancetherapy.However,thepublisheddataareconflicting(Table2).Somestudiesreportednobenefitofmaintenancetherapyincontrollingtumourrecur-rence,12–14,16whereasasignificantdifferencewasseeninrecur-rence-freesurvivalinonestudyfavouringmaintenancetherapy.15Thenegativestudiesrecruitedfewerpatientsandwereunderpoweredstatistically.Themaintenanceschedulesvariedconsiderablyanditappearsthatamonthly,12quarterly13or6monthlyschedule16foronetotwoyearmaybeimmunologicallysuboptimal.Thepositivestudy15recruitedalargernumberofpa-tientsandcontinuedmaintenancetherapyfor3years.Significantimprovementswereobservedintumourrecurrenceanddiseaseprogressionwithasuggestionofasurvivaladvantage.Meta-anal-yseshaveshownthatBCGmaintenancesignificantlyreducesrecurrenceinpatientswithsuperficialbladdercancer8and,impor-lavivM%%35r+)Rus)N8ll0ea.uree0lavMRRRt%RandomisedtrialsofintravesicalBCGandMitomycinC(MMC).StudyPatientsInterventionsRecurrenceProgressionOverallsurvivalBCGMMCBCGMMCBCGMMCRintala21109RecurrentTa/T1G1–3,75mgPasteurBCG(6Â108CFU)CRIncidenceNRorCISweeklyÂ4,monthlyÂ24versus97%70%9.8%5.2%MMC20–40mgscheduleasaboveDFS:BCGsignificantlylonger(p=0.001)(nopvalue)Lamm22469RecurrentTa/T1.50mgTICEBCG(5Â108CFU)RFSsignificantlylongerwithBCG(p=0.017)NodifferenceinprogressionRateorconcurrentCISweeklyÂ6,8and12weeks,(p=0.48)monthlyÂ12,versus20mgMMC.5%88%scheduleasabove(p=0.57)TTDNS(p=0.57)Krege23337pTa/pT1G1–3120mgConnaughtBCG1h+0.5mgIncidence%peryearNR(nopTaG1tumours)BCGs.c.weeklyÂ6,monthlyÂ425%27%4.532.01versus20mgMMC2weeklyÂ26,(logrankp=0.255)(p=0.207)monthlyÂ12Witjes24469pTa/pT1orCISBCGRIVM(5Â108CFU)weeklyÂ6IncidenceIncidenceNRversus80mgTICEBCGasabove46%(RIVM)6%(RIVM)versusMMC30mgweeklyÂ4,%(TICE)5%(TICE)monthlyÂ6.43%6%(MMCvsRIVMNS)Witjes25344PrimaryorrecurrentBCGRIVM5Â108CFUweeklyÂ6TTR:nosignificantdifference(logrankp=0.376)Nosignificantdifference(logrankRatepTa/pT1,CISversus30mgMMCweeklyÂ4,p=0.09)monthlyÂ68.7%10.4%(NS)Malstrom26126MultipleorrecurrentBCGPasteur120mg(109CFU)Disease-freeat5years:IncidenceRateat5yearsTa/T1G1–2,CIS,T1G3weeklyÂ6,monthlyÂ12,3-47%34%16%21%75%80%monthlyÂ8versusMMC40mgas(p=0.04)(p=0.39)(p=0.90)aboveMangiarotti2796LowgraderecurrentBCGTICE(ndosegiven)weeklyÂ6,%Recurrence-freeNoprogressionsobservedNRTa/T1monthlyÂ12versusMMC40mg50%50%weeklyÂ6,monthlyÂ8TTR(months):21.917.5Ojea28430Ta/T1G1–2ConnaughtBCG27mgweeklyÂ6,DFI:BCG27mgsignificantlylongerthanMMCIncidenceNosignificantdifferenceincancerbi-weeklyÂ6versusBCG13.5mgas(p=0.006)specificsurvivalaboveversusMMC30mgasaboveBCG27mg10%(BCG27mg2%,BCG13.5mg4%,MMC5%)BCG13.5mg13%9%(notsignificant)Sekine20012942consecutivepatientwithBCGTokyo80mgweeklyx6versusCRIncidenceCancerdeathsprimaryorsecondaryCISMMC20mgday1andADR40mgfor86%81%10%5%10%5%5consecutiveweeks(p=0.68)(NS)(NS)Rate11%52%(p=0.003)Freidrich32495Recurrentand/orRIVMBCG(2Â108CFU)weeklyÂ6,RFat3yearsNRNRmultiplepTaG1,versus20mgMMCweeklyÂ6,65%TaG2–3,T1G3versus20mgMMCweeklyÂ6,69%monthlyÂ36.86%DiStasi34212pT1.AllhadTUR81mgBCG2hinstillationweeklyÂ6DFIIncidenceIncidenceOverallmortalityandmultiplebiopsies.versusBCGweeklyÂ2plusMMC2158%22%32%40mgelectromotively(20mAfor6942%9%%22%30min)weeklyÂ3p=0.001p=0.001p=0.005p=0.004BCG–bacillusCalmette–Guerin,CIS–carcinomainsitu,CFU–colonyformingunits,CR–completeresponse,NR–notreported,DFS–disease-freesurvival,DFI–disease-freeinterval(months),RFS–recurrence-freesurvival,TTD–timetodeath,NS–notsignificant,s.c.–subcutaneous,TTR–timetorecurrence,RF–recurrence-free.198M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205199tantly,inT1G3tumourswhichareathighriskofrecurrenceandprogression.17TheprolongedadministrationofBCGinmaintenancetherapymaylogicallybeassociatedwithincreasedtoxicitycomparedtonomaintenance.However,areportbytheEORTCGenitourinaryGroup18on487patientsreceivinga3yearBCGmaintenancesche-duleaspartofarandomisedstudy,observedthatthemajorityofside-effectswereseenduringtheinductionperiodandthefirst6monthsofmaintenance,suggestingthatfurthermaintenanceisnotproblematic.Inaddition,ameta-analysisreportednosignifi-cantdifferenceintheincidenceofcystitisinpatientsreceivingmaintenanceBGCcomparedtothosewhoreceivedinductionther-apyalone.3ItisnowgenerallyacceptedthatsomeformofBCGmaintenancetherapyisbeneficialbuttheoptimalschedulere-mainstobedetermined.Thisisimportantastherearecostandmorbidityimplicationswithmaintenancetherapy.Thereisalsoaneedtodeterminewhichpatientswouldbenefitmostsothatthosewhoremaineddisease-freeafterinductiontherapywouldnotbeover-treated.SinceanessentialimmunologicalcomponentofthemechanismofintravesicalBCGappearstobetumourinfiltrationofT-cells,at-temptshavebeenmadetobooststimulationbyconcomitantper-cutaneousandintradermalBCGadministration.However,tworandomisedtrialshaveshownthatsimultaneouspercutaneous19orintradermal20BCGdoesnotsubstantiallyimprovethethera-peuticeffectofintravesicalBCG.RandomisedtrialscomparingintravesicalBCGwithintravesicalchemotherapyBCGversusMitomycinCNinepublishedrandomisedtrialshavedirectlycomparedintra-vesicalBCGwithintravesicalMitomycinC(Table3).21–29Thedosesandschedulesforeachinterventionvariedconsiderably.IntermsTable4

RandomisedtrialsofalternatingorcombiningintravesicalBCGandMitomycinC(MMC).StudyRintala36ofrecurrence,fourstudiesreportedthatBCGwassignificantlysuperiortotheMMCschedulesused.21,22,26,28Inthreeofthesetri-alsthepatientsrecruitedwereconsideredhighriskfortumourrecurrence.21,22,26Nosignificantdifferencewasnotedintheremainingstudieswhichrecruitedintermediateorintermediate–highriskpatients.Ameta-analysiscomparingBCGwithMMCsug-gestedthatBCGwassuperiorfortheprophylaxisoftumourrecur-renceinhighriskpatientsonly.30Similarly,themeta-analysisconductedbyBohleetal.3indicatednosignificantdifferenceinrecurrenceratesbetweenBCGandMMCforintermediateriskpa-tients.TheevidencesuggeststhatthereisnoclearadvantageofBCGoverMitomycinCinpatientswithintermediateriskofrecur-rencealthoughtoxicitymightbemoresevere.ChemotherapymightthereforebeanalternativetoBCGinthisriskgroup.How-ever,maintenancetherapywithBCGisrequiredtoclearlydemon-stratethesuperiorityofBCG.31,3Long-termMMCappearstobesuperiortoshort-termBCG,howeverthelatterisnotroutineclin-icalpractice.32Therandomisedtrialsreportednosignificantdiffer-enceindiseaseprogressionorsurvivalbetweenBCGandMMC,althoughonemeta-analysis,reportedthatBCGwasmoreactivethanMMCinthepreventiontumourprogression.33However,are-centindividual-patient-datameta-analysiswhichhadalongerfol-lowthanpreviouslypublishedmeta-analyses,indicatedthattherewasnosignificantdifferenceintermsofdiseaseprogression.31AllstudiesthatreportedtoxicityindicatedthatitwasmoreseverewithBCG.AllthisevidencetendstosuggestthatintravesicalBCGwithmaintenance,althoughmoretoxic,hassuperioractivitycomparedtoMMCfortheprophylaxisoftumourrecurrenceinhighriskpa-tientsalthoughthereisnosignificantdifferenceindiseaseprogres-sionandoverallsurvival.ChemotherapyandBCGmaybeequallyeffectiveinpreventingrecurrenceinintermediateriskgroups.EncouragingdatahavebeenreportedusingacombinationofBCGpluselectromotivelyadministeredMMCcomparedtoBCGalone.34ThosepatientsrandomisedtothecombinationhadaPatients188RecurrentpTa/pT1InterventionsAllreceivedMMC20–40mgweeklyÂ5thenrandomisedto:MMC20–40mgversusalternatingMMCand75mgPasteurBCG(6Â108CFU).BotharmsmonthlyÂ12,3-monthlyÂ3RecurrenceTTR(months)1.01120.867(p=0.38)(p=0.97)CRRI70%79%88%11%(p=0.06)RecurrencesignificantlybetterwithBCGmono-therapy(disease-free41%versus55%,p=0.02)Incidence46%40%(p=0.36)DFSsignificantlyprolongedwithMMCinduction(p=0.03)MonthlyBCGsuperior.TTR(logrankp<0.00001).Rate(0.4versus0.9,p<0.0001).RI(0.9versus0.3,p<0.0001)RatesProgressionIncidence3.2%3.2%Incidence18%7%(p=0.035)ProgressionrisksignificantlylowerwithBCGmono-therapy(p=0.07)OverallsurvivalNRG-Mas3766RecurrentpTa/pT1G1–3MMC40mgmonthlyÂ12versusalternatingMMCandBCG54mgmonthlyÂ12NRKaasinen38323Primary,secondaryorconcurrentCISMMC40mgweeklyÂ6,thenalternatingBCG(120mg)andMMCmonthlyÂ12versusBCG120mgweeklyÂ6,monthlyÂ12Nosignificantdifference(nopvaluegiven)Witjes39Solsona40182RecurrentormultiplepTaorpT1G3,primaryorconcurrentCIS405MultipleorrecurrentTaG2–3,T1G1–3,CIS236RecurrentTa/T1MMC40mgweeklyÂ4thenTICEBCG(5Â108CFU)weeklyÂ6versusMMC40mgweeklyÂ10Incidence4.3%5.5%(p=0.70)Nosignificantdifference(nopvaluegiven)NRMMC10–30mg24hbeforeBCG(Connaught81mg)weeklyÂ6,thebi-weeklyversusBCG(samescheduleasabove)AllreceivedMMC40mgÂ5thenrandomisedto:BCG(Oncotice5Â108CFU)monthlyÂ12versusalternatingmonthlyINF-a2b(50mU)andBCGÂ12Nosignificantdifference(nopvaluegiven)NRKaasinen41NRBCG–BacillusCalmette–Guerin,CIS–carcinomainsitu,CFU–colonyformingunits,TTR–timetorecurrence,CR–completeresponse,IFN–interferon,RI–recurrenceindex.200M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205significantlylongerdisease-freeinterval,andsignificantlyreducedratesforrecurrence,progressionandoverallmortality.Ameta-analysiscomparinglongandshort-termintravesicalBCGwithintravesicalchemotherapy(MitomycinC,EpirubicinandAdriamy-cin)inpatientswithcarcinomainsitu,demonstratedthesuperior-ityofBCGintermsofcompleteresponseanddisease-freesurvival.35AlternatingorcombiningBCGandMitomycinCAnumberofrandomisedtrialshaveattemptedtoreducetheharmfuleffectsofBCGandimproveprophylaxisforrecurrencebycombiningoralternatingMMCwithBCG(Table4).36–41ThenumberofpermutationsofMMCandBCGisconsiderable,conflict-ingandconfusing.Inpapillarytumours,alternatingMMCwithBCGwasshowntobeequivalenttoMMCmono-therapyinoneran-domisedtrial36butsuperiorinanother.37Whenshort-termMMCplusBCGwascomparedtolong-termMMCtherewasnosig-nificantdifferenceinrecurrenceorprogressioninhighriskpa-tients.39However,MMCplusBCGsignificantlyimproveddisease-freesurvivalcomparedtoBCGmono-therapybutagaintherewasnodifferenceinprogressionorsurvivalrates.40Inaddition,thiscombinationwasmoretoxicthanBCGmono-therapy.Finally,followinginductionwithMMC,nobenefitwasobservedbyalter-natinginterferonwithBCGcomparedtoBCGalone.41ItisdifficulttomakeanyrationalconclusionfromthesedatabutnoneofthecombinationswereshowntobesuperiortoBCGaloneintermsofprogressionandsurvivalandsomeweremoretoxicthanBCG.BCGversusAdriamycinTheanthracyclineantibiotic,Adriamycin(Doxorubicin),hasestablishedactivityinanumberofcancersincludingsuperficialbladdercancer.Toascertaintherelativeclinicalbenefitofintraves-icalAdriamycinrandomisedstudieshavecomparedthisagentwithintravesicalBCG(Table5).42–45ThetrialresultsconsistentlyreportthatBCGissuperiortoAdriamycininpreventingordelayingtumourrecurrencebothinpapillaryTaandT1cancersandCIS.ThereissomeevidencetosuggestlowerprogressionrateswithBCGbutnodifferenceinoverallsurvival.43,44BCGversusEpirubicinIntravesicalEpirubicin(40epidoxorubicin)hasshownactivityintheprophylaxisoftumourrecurrenceinTaandT1superficialblad-dertumoursalthoughitseffectonprogressionislessclear.Anum-berofrandomisedtrials(Table6)havecomparedtheefficacyofintravesicalBCGwithEpirubicintodeterminetheoptimumintra-vesicalagentandschedule.46–50IntravesicalBCGsignificantlyprolongedthetimetofirstrecur-rence46,47comparedtoEpirubicinandimprovedtherecurrencerate.TherewasnosignificantdifferencebetweenBCGandEpirubi-cinintermsofdiseaseprogressionoroverallsurvival.BCGinducedside-effectswerenotamelioratedbyalternatingBCGandEpirubi-cin48bysequentialadministration50orbyconcomitantisoniazidwithBCG.49BCGversusotheragentsRandomisedstudieshavereportedthesuperiorityofintravesi-calBCGcomparedtointravesicalthiotepainreducingtumourrecurrenceinpatientswithsuperficialbladdercancer.44,51Thecon-comitantuseoforaltegafurwithBCGdoesnotsignificantlyim-proverecurrenceorprogressionratescomparedtoBCGalone.IntravesicalBCGconsistentlyappearstoinducemoretoxicitythanintravesicalchemotherapy.InanattempttoimproveBCGtol-eranceonestudyrandomised116patientstoantbioprophylaxiswithoralofloxacineorplacebofollowinginstillationsofBCG.52Re-sponseratesweresimilarinbotharms,however,severeadverseevents,whichgenerallydeterminepatient’swithdrawalfromtreatment,weresignificantlyreducedwithofloxacine.Alongerfol-Table5

RandomisedtrialsofintravesicalBCGandAdriamycin(ADR).StudyPatientsInterventionsRecurrenceBCGKhanna42ProgressionADR66%22%CombinedrecurrenceandprogressionratesforCIS42%65%BCGIncidence1%ADR0%OverallsurvivalBCGNRADR116RecurrentTa,T1andCIS(initiallyrandomisedthenopenlabel)Lamm43285RecurrentTa/T1/CISTICEBCG2hversus50mgADR2h.Bothschedules:weeklyÂ6,monthlyÂ12,3-monthlyÂ8,6-monthlyÂ4120mgConnaughtBCG2hweeklyÂ6,3-monthlyÂ3,6-monthlyÂ2versusADR50mg/0.5hweeklyÂ4,monthlyÂ11.CR83%Rates3%CRRates.6%(NS)61.6%Martinez-Pineiro44202PatientswithpTa/pT1150mgPasteurBCGversus50mgADRversus50mgThiotepa(TP).Allinstillations1h:weeklyÂ4,monthlyÂ11.(noBCGdosegiven)80mgTokyo172BCGweeklyÂ6versusADR20mgweeklyÂ2,2-weeklyÂ7,monthlyÂ8.Hinotsu4583pTa/pT1G1–370%34%(p<0.001forCIS)MedianTTF(months)375.1(p<0.001)BCGADRTPRate(%)134336(p=0.002BCGversusADR,p=0.04BCGversusTP)DFS:significantlybetterwithBCG(p<0.001BCGversusADR,p<0.01BCGversusTP)BCGreducedriskofrecurrence(logrankp=0.017)BCGADRRate(%)1.57.5(nopvaluesreported)TP3.6NRNRNRNS–notsignificant,BCG–BacillusCalmette–Guerin,CIS–carcinomainsitu,CR–completeresponse,NR–notreported,NS–notsignificant,DFS–disease-freesurvival,TTF–timetofailure.M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205Table6

RandomisedtrialsofintravesicalBCGandEpirubicin(EP).StudyPatientsInterventionsRecurrenceBCGCheng46201ProgressionEPBCGEPOverallsurvivalBCGEP209Patientswithstage>Ta,grade>1,>1cm,ormultipleorrecurrenttumours47deReijkeConnaughtBCG81mg(6.6–19.2CFU)2hinstillationweeklyÂ6,monthlyÂ10versus50mgEP2hinstillationweeklyÂ4,monthlyÂ5,3monthlyÂ2168Patientswithprimary,ConnaughtBCG81mgweeklyÂ6secondaryorconcurrentCISversus50mgEPweeklyÂ8.(3yearsmaintenancetherapyforpatientswithCRinbothgroups)Ali-el-Dien48VandeMeijden49139Patientswithrecurrent150mgPasteurBCG(5Â108–5Â109CFU)versusalternatingBCGandgrade2–3pTaandpT150mgEP(allinstillations(2h)givenweeklyÂ6themonthlyÂ10)957PatientswithTa–T1TICEBCG5Â108CFUweeklyÂ6DFIfavoursBCGcomparedtoGrade1–3completelyversusBCG+300mgisoniazidversusEP:BCGalone(p=0.001),resectabletumours50mgEPweeklyÂ6(allgroupshadBCG+isoniazid(p=0.009)maintenancetherapyfor3years)Incidence29%55%(p=0.004logrank)10yearRFS61%32%MedianTTR5.1years1.4years(p=0.0004)CISrecurrence45%16%RFSlongerwithBCG+EP(p=0.05)10yearPFS10yearOSrate78%74%74%53%(p=0.66logrank)(p=0.57)10yearDSSrate92%53%IncidenceRate27%18%69%60%(p=0.17)(p=0.26)NSNRIncidence3%BCG/isoniazid5%7%Bilen20005041PatientswithpT1(either81mgConnaughtBCGweeklyÂ6grade3ormultipleorversussequentialBCGand50mgEPrecurrentor>4cmweeklyÂ12tumours)Incidence19%(p>0.05)15%Incidence10%(p>0.05)Medianfollow-up3.9years.Nosignificantdifferencebetweengroups(p=0.26)NR5%BCG–BacillusCalmette–Guerin,CFU–colonyformingunits,RFS–recurrence-freesurvival,PFS–progression-freesurvival,DSS–disease-specificsurvival,CIS–carcinomainsitu,CR–completeresponse,TTR–timetorecurrence,RR–relativerisk,DFI–disease-freeinterval,NS–notsignificant.,NR–notreported.,TUR–transurethralresection.Table7

RandomisedtrialsofintravesicalBCGversusotherintravesicalimmunomodulators.StudyJimenez-Cruz54Participants122RecurrentpT1G1–3tumoursInterventionsBCGPasteur150mgweeklyÂ4,bi-weeklyÂ8,monthlyÂ9versusIFNa2a54MUasaboveBCGConnaught120mgweeklyÂ5,monthlyÂ4,3-monthlyÂ2versusKLH1mgintracutaneouslythen10mgintravesicallyasaboveFulldoseBCG150mgPasteur6weeklyinstillationsversusLowdoseBCG75mgplusIFN2b10MUAllhadpirarubicin40mgintravesicallycontinuously3dayspostTURthenrandomisedto:BCGTokyostrain40mgweeklyÂ6versusBCGplusIL21KUweeklyÂ6RecurrenceIncidence39.3%69.4%p=0.009Incidence14%41.2%DFI19.315.3p=0.009Rate0.761.95ProgressionIncidence9.8%14.3%p=0.673Incidence0%11.8%NRKalble5542Ta/T1,CISG1–3recurrentormultifocaltumoursBercovich5636RecurrentTa/T1G1–3Ao5740pTa,pT1a,pT1b(excludedCIS)Incidence27.8%22.2%,p=0.84Recurrence-freeratesat2years.5%.9%,p=0.48NRCIS–carcinomainsitu,IFN–interferon,DFI–disease-freeinterval(months),IL2–interleukin-2,NR–notreported,KLH–key-holelimpethaemocyanin.low-upisrequiredtomoreaccuratelydeterminetheimprovementintheclinicalbenefitprofileofBCGwithofloxacine.AnotherendeavourtoreducethefrequencyandseverityofBCGside-effectswasreportedinasmallgroupofuntreatedpatientswithCIS(n=55)randomisedtointravesicalBCGortheimmuno-modulatororalbrompirimine.53Althoughtherewerenosignificantdifferencesinresponserates,thedrop-outrate,irritativeurinarycomplaintsandhaematuriaweresignificantlylesswithbrompiri-mine.Nofurtherrandomisedstudiesofthisagenthavebeenpub-lishedsincethisstudyandmayreflecttheside-effectsassociatedwiththisagent.BCGversusotherimmunomodulatorsAlthoughintravesicalBCGhasbeenusedformanyyearstotreatsuperficialbladdercanceritsmechanismofactionremainstobeclarified.However,itisknownthatfollowingBCGtreatmentanumberofcytokinesaredetectableintheurineincludinginter-feron-c(IFN)andinterleukin-2whichprobablyreflectslocalacti-vationofT-cells.ThishasledsomeresearcherstoinvestigatetheprophylacticpotentialofotherimmunomodulatorsapartfromBCG(Table7).ResultsfromrandomisedstudiesindicatethatBCGhassignificantlybetteractivityinreducingrecurrencethaneitherIFNorkey-holelimpethaemocyanin.54,55TheadditionofIL-2toBCGdoesnotenhanceits’efficacy.56,57Theevidencefromrandom-isedtrialsindicatesthatintravesicalBCGappearstobethemostactiveimmunomodulatoravailableatpresent.Mitomycin–comparisonofschedules,otherintravesicalagentsandproceduresAnumberofrandomisedtrialshaveattemptedtodefinetheoptimumscheduleforMitomycinC(Table8).Thedatasuggestthatasignificantreductionintumourrecurrencemaybeachievedwhenusing40mgcomparedto30mg,aonehourinstillationcom-paredto0.5h,and5instillationsratherthanjustone.58,59An202M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205intensiveweeklyinductionwithmaintenancetherapyappearstobesuperiortoamildinductionwithminimalmaintenance.60,61MitomycinChasbeencomparedtoaseveralotherintravesicalagents(Table8).Itissignificantlymoreeffectivethanthiotepainpreventingordelayingtumourrecurrence,althoughthereisevi-dencethatMitomycinismoretoxic.62TheevidencealsosuggeststhatMitomycinCissuperiortointravesicalinterferonintermsoftumourrecurrence,althoughinterferonhasaslightlybettersafetyprofile.63MitomycinChasshownhigherrecurrence-freeratesandlowerprogressionratesthanintravesicalAdriamycinbutthedifferenceshavenotachievedstatisticalsignificance.,65TreatmentwithintravesicalGemcitabinehasdemonstratedsignif-icantlybetterprophylaxisforrecurrencethanMitomycinCbutwithnosignificanteffectonprogression.66Severalnovelprocedureshavebeenusedtoimprovetheeffi-cacyofintravesicalMitomycinC.Onegroupofresearchershasuti-lisedpharmacokineticmanipulation,wherebyoralsodiumbicarbonatewasusedtoalkalinisetheurineincreasingthestabilityofthedrug,whilstreducingtheurinevolumepriortoinstillation,thusincreasingthebladderdrugconcentration.67Thosepatientsrandomisedtothepharmacokineticmanipulationarmhadfewerrecurrencesandasignificantlylongertimetorecurrence.Combin-ingMitomycinCwithlocalmicrowavehyperthermiatoheatthebladderwall68orapplyinganelectromotivecurrenttoimprovedruguptake69haveshownsignificantlyimprovedrecurrenceparameterscomparedtoMitomycinCalone.AproposedHYMNrandomisedtrial(ASRCTN85785327)aimstodeterminewhetherhyperthermiaplusintravesicalMitomycinCiseffectiveinpatientswithrecurrenturothelialcellcarcinomafollowingintravesicalBCGinductionormaintenancetherapy.TheproposedstartingdateforthisstudyisJune2009;itwillthereforebesometimebeforetheresultsofthisstudyarepublished.Howeveratpresent,limitedexpertiseinthesemethodsisavailabletoapplythemforroutineuse.Table8

RandomisedtrialsofMitomycinC(MMC):comparisonofschedules,otherintravesicalagents,andprocedures.StudySchedulesScaldazzaTolley5958Participants60pTaorpT1G1–3502Ta/T1G1–3InterventionsAllhadTUR.MMC30mgfor1yearversusMMC40mgfor6monthsTURaloneversusMMC40mgÂ1versusMMC40mgÂ5(3monthly)MMC40mgweeklyÂ8monthlyÂ4versusweeklyÂ4monthlyÂ5versusweeklyÂ8monthlyÂ12versusweeklyÂ4monthlyÂ12MMC20mg3-yeartherapyversusintensive3-yeartherapyversusintensiveshort-termtherapyversusAdriamycin50mg3-yeartherapyRecurrenceIncidence:40%(rate3.3)27%(rate2.2.)Nodifferencebetweengroups.CombiningMMCgroupsversusTURHR0.84(95%CI0,.45–1.35,p=0.56)Recurrence-freeintervallongerwithintensivestartregimeandmonthlyÂ12(maintenance).Incidence24.4%17.7%20.0%30.8%CR39%26%(p=0.01)Responserates72%19%33%41%Recurrence-freerates32%25%7%Recurrences/year0.470.470.510.51Recurrence-free67%78%(p=0.05)ProgressionNRNRNewling601282MultipleorrecurrenceTaorT1tumoursNRSchaibold61419PrimaryorrecurrentTa,T1CIS,G1–3tumoursIncidence11.5.%5.2.%6.7%17.9%NROtheragentsHeney62176ResidualTa/T1tumours(>2cm)115MultipleprimaryorrecurrentTa–T1withmarkerlesionAllhadTUR.MMC40mgweeklyÂ8versusThiotepa30mgweeklyÂ8AllhadTUR.MMC40mgweeklyÂ8versusIFN30MUweeklyÂ12versusIFN50MUasaboveversusIFN80MUasaboveMMC30mgÂ6in2weeks,monthlyÂ24versusADR50mgasaboveversusTURaloneMMC20mgweeklyÂ2,bi-weeklyÂ7,monthlyÂ8,3-monthlyÂ4versusADR30mgasaboveversusADR20mgasaboveversusTURaloneMMC(nodosegiven)weeklyÂ4(maintenancemonthlyÂ10forthoserecurrence-free)versusGemcitabine(nodosegiven)weeklyÂ0mgMMCplusPKmanipulationversus20mgMMC(bothweeklyÂ6)Malstrom63NRTsushima110RecurrentTa–T1tumoursNRAkaza65158pTa,pT1,CIS,G1–3singleormultipleFaiola66120RecurrentTa/T1G1–3Incidence27%43%31%39%Incidence10.8%11.1%(NS)ProceduresAu67230pTa/pT1,CIS,G1–3singleormultifocal,primaryorrecurrent83PrimaryorrecurrentormultipleTa/T1G1–3Colombo68MMC20mgpluslocalmicrowavehyperthermia(42°Cforatleast40min)versusMMC20mg40mgElectromotiveMMC(20mAcurrent30min)versus40mgpassiveMMCversus81mgBCGPasteur108CFU(allweeklyÂ6,monthlyÂ10).DiStassi69117withCISmostwithconcurrentT1tumoursTTR29.011.8(p=0.005)Incidence17%58%(p=0.0002)CR58%31%56%Recurrence-free41%24.6%TTR:earlierwithMMCalone(p=0.0002)NROnly1reportedinhyperthermiagroupTTR35.019.526.0NRTUR–transurethralresection,NR–notreported,RI–recurrenceinterval(months),CI–confidenceinterval,NS–notsignificant,CIS–carcinomainsitu,CR–completeresponse,INFa–interferonalpha,TTR–timetorecurrence(months),RRR–relativerecurrencerate,ADRAdriamycin,PK–phamacokinetic.M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205203ImmediatesingleinstillationofchemotherapyOneofthemechanismscommonlyacceptedforsuperficialbladdercancerrecurrenceisthere-implantationofcirculatingtu-mourcellsreleasedduringthesurgicalprocessofTUR.ThishadledsomeworkerstoinvestigatethepotentialofinstillingcytotoxicagentsintothebladderimmediatelyafterTURwiththeaimofkill-inganyfreeintravesicaltumourcellsandreducingrecurrencerates.Forexample,arandomisedstudyhasshownthatoneinstil-lationofMMCimmediatelyafterTURreducestherateofrecur-renceby34%whereas5instillationsreduceitby50%.59Theside-effectscommonlyassociatedwiththisagenttendtobechemicalcystitisandcontactdermatitis.SimilarlyinanotherrandomisedtrialofpatientswithTaandT1tumours,animmediatesingleinstillationofEpirubicin,hasshowntoreducetherateofrecur-rencebyhalfcomparedtoaninstillationofwater.70ThishasbeenconfirmedinamorerecentstudywherethecontrolgroupreceivednofurthertreatmentfollowingTURbuthighlightedthatthebene-fitwasmostprofoundinpatientswithprimary,solitarytumoursandminimalforrecurrentormultipletumours.71Inarecentran-domisedplacebo-controlledtrial,theefficacyofasinglepost-oper-ativeinstillationofGemcitabine(2g/100mlofsaline)wasreported.72Inthisstudyof355patientswithprimaryorrecurrentTa/T1G1-3tumours,Gemcitabinewasnotsuperiortoplacebo(sal-ine)intermsofrecurrence-freesurvival.Toclarifythisfurther,ameta-analysiswasperformedincluding7randomisedtrials,eachevaluatingtheclinicalvalueofasingleimmediateinstillationofacytotoxic,andreporteda39%decreaseintheoddsofrecurrencewithchemotherapy.73Bothpatientswithsingleandmultipletumoursbenefited,however,thosewithmulti-pletumourshadsignificantlymorerecurrences.Toxicitywasmildandtransient,mainlydysuria,frequencyandhaematuriainabout10%ofpatients.Thesedatasuggestthatasingleimmediateinstil-lationofchemotherapymaybethebestinitialoptionforbothpa-tientsofallriskcategoriesbutthosewithhighriskstumourswouldrequireadditionaltherapy.However,incaseswherebladderperforationissuspectedorincasesofgrosshaematuria,animme-diateinstillationshouldnotbeadministeredasthismayresultinchemicalperi-cystitis.Althoughthisconditionisrareitdoespres-entwithlifethreateningsymptomsandshouldbenotedwhenconsideringhighriskpatientsforimmediateinstillationtherapy.ConclusionsTheanatomicallocationofsuperficialbladdercancerishighlyconducivetoregionaltherapy.Thebladderisanorganwithawelldefinedcavitythatallowstheinstillationoftherapeuticagentsthatcanbathethetumoursathighconcentrationswhilstlimitingsys-temicexposureduetotheimpermeablenatureofthebladderwall.Eliminationoftheagentiseasilyachievedbyflushingandmicturition.Withthesecharacteristicsitisnotsurprisingthatadjuvantintravesicaltherapyhasbeenwidelyexploredasatreatmentop-tionforsuperficialbladdercancer.Infactthenumberofrandom-isedtrialsisextensiveandduetothejournallimitationonspace,manyolderrandomisedtrialshavenotbeencitedinthisreview.Whatisapparentfromtheincludedstudies(Tables2–8)isthattheinterventionsinvestigated,bothsinglyandincombination,arewide-ranging.Inaddition,thereisconsiderablevariationinthetimebetweenTURandtreatment,dosagesused,exposuretimes,frequencyanddurationoftreatments.Thecomplexityofinterventioncomparisonsillustratedinthissystematicreviewmakesitdifficulttoclearlyassesstheavailableevidence.Itwouldbeasensibleapproachtodefineagoldstandardintra-vesicalregimeandcompareallotherintravesicalregimeswiththis.Thereisgoodqualityevidencethatasingleinstillationofche-motherapywithin24hofTURhasasignificanteffectinreducingtumourrecurrenceandcouldformtheinitialstepinthe‘goldstan-dard’treatment.Thismaybeallthatisrequiredforlowrisktu-moursbutintermediateandhighrisktumourswouldrequireadditionaltherapy.ThemajorityoftheevidencesuggeststhatintravesicalBCGwithmaintenancetherapyshouldbeofferedtothesepatientsasthesecondstepofthe‘goldstandard’regime.Chemotherapy,withMitomycinCorEpirubicin,wouldbeanop-tionforthosefailingorwhoareunsuitableforBCGtherapy.EventhoughBCGhasbeenusedforsomeconsiderabletime,theoptimumdoseandschedulehasyettobedetermined.Thisisavi-talareaoffutureresearchastheseimpacts,notonlyonclinicaleffectiveness,butalsoonpatientmorbidity,qualityoflife,andmedicalcosts.AttemptstoreduceBCGtoxicitylookpromisingandshouldalsobeamajorfocusforfuturetrials.ConflictofintereststatementNonedeclared.AcknowledgementsTheauthorswouldliketothankMrs.BernadetteColesfordevelopingandrunningthesearchstrategyforthisreviewandCancerResearchWalesforfundinglibraryfacilities.References1.ParkinD.Theglobalburdenofurinarybladdercancer.ScandJUrolNephrol2008;218(Suppl.):12–20.2.MoralesA,EidingerD,BruceC.IntracavitybacillusCalmette–Guerininthetreatmentofsuperficialbladdercancer.JUrol1976;116:180–3.3.BohleA,JochamD,BockPR.IntravesicalbacillusCalmette–GuerinversusmitomycinCforsuperficialbladdercancer:aformalmeta-analysisofcomparativestudiesonrecurrenceandtoxicity.JUrol2003;169(1):90–955.4.SylvesterRJ,vanderMeijdenAP,LammDL.IntravesicalbacillusCalmette–Guerinreducestheriskofprogressioninpatientswithsuperficialbladdercancer:ameta-analysisofthepublishedresultsofrandomizedclinicaltrials.JUrol2002;168(5):19–70.5.HuncharekM,GeschwindJF,WitherspoonB,McGarryR,AdcockD.Intravesicalchemotherapyprophylaxisinprimarysuperficialbladdercancer:ameta-analysisof3703patientsfrom11randomizedtrials.JClinEpidemiol2000;53(7):676–80.6.PawinskiA,SylvesterR,BouffiouxC,KurthKH,ParmarM,BijnensL.AcombinedanalysisofEORTC/MRCrandomizedclinicaltrialsfortheprophylactictreatmentofTaT1bladdercancer.EortcGenito-UrinaryTractCancerCooperativeGroupandtheMedicalResearchCouncilWorkingPartyonsuperficialbladdercancer.ActaUrologicaBelgica1996;(2):27.7.ShelleyMD,CourtJ,BurgonK,etal.Meta-analysisofintravesicaltherapyforsuperficialbladdercancer:superiorityofbacillusCalmette–Guerinmaybeconfinedtohighriskpatients.BrJCancer2001;85:53.8.HanRF,PanJG.CanintravesicalbacillusCalmette–Guerinreducerecurrenceinpatientswithsuperficialbladdercancer?Ameta-analysisofrandomizedtrials.Urology2006;67(6):1216–23.9.Martinez-PineiroJ,FloresAN,IsornaS,etal.Long-termfollow-upofarandomizedprospectivetrialcomparingastandard81mgdoseofintravesicalbacilleCalmette–Guerinwithareduceddoseof27mginsuperficialbladdercancer.BJUInt2002;(7):671–80.10.Martinez-PineiroJ,Martinez-PineiroL,SolsonaE,etal.ComparisonofastandardBCGdose(81mg)versusathree-foldreduceddose(27mg)inhighrisksuperficialbladdercancer(T1G3,Tis).ACUETOprospectiverandomisedstudy95012.EurUrol2003;2(Suppl.1):190.11.BohleA.Amulticentre,randomisedprospectivetrialcomparingthreeintravesicaladjuvanttherapiesforintermediate-risksuperficialbladdercancer:low-dosebacillusCalmette–Guerin(27mg)versusverylow-dosebacillusCalmette–Guerin(13.5mg)versusmitomycinC.IntBrazJUrol2008;34(1):117–8.12.BadalamentRA,HerrHW,WongGY,etal.AprospectiverandomizedtrialofmaintenanceversusnonmaintenanceintravesicalbacillusCalmette–Guerintherapyofsuperficialbladdercancer.JClinOncol1987;5(3):441–9.13.HudsonMA,RatliffTL,GillenDP,HaaffEO,DresnerSM,CatalonaWJ.SinglecourseversusmaintenancebacillusCalmette–Guerintherapyforsuperficialbladdertumors:aprospective,randomizedtrial.JUrol1987;138(2):295–8.14.AkazaH,HinotsuS,AsoY,KakizoeT,KoisoK.BacillusCalmette–Guerintreatmentofexistingpapillarybladdercancerandcarcinomainsituofthebladder:four-yearresults.Cancer1995;75(2):552–9.204M.D.Shelleyetal./CancerTreatmentReviews36(2010)195–20515.LammDL,BlumensteinBA,CrawfordED,etal.MaintenancebacillusCalmette–GuerinimmunotherapyforrecurrentTA,T1andcarcinomainsitutransitionalcellcarcinomaofthebladder:arandomizedSouthwestOncologyGroupStudy.JUrol2000;163(4):1124–9.16.PalouJ,LagunaP,Millan-RodriguezF,HallR,Salvador-BayarriJ,Vicente-RodriguezJ.ControlgroupandmaintenancetreatmentwithbacillusCalmette–Guerinforcarcinomainsituand/orhighgradebladdertumors.JUrol2001;165(5):1488–91.17.PanJ,ZhouX,HanR,ChenZ.Ameta-analysisofrandomizedtrialsofmaintenancebacillusCalmette–GuerininstillationefficacyagainstrecurrenceofT1G3bladdertumor.FrontMedChina2008;2(3):259–63.18.vanderMeijdenAP,SylvesterR,OosterlinckW,HoeltlW,BonoAV.MaintenancebacillusCalmette–GuerinforTaT1bladdertumoursisnotassociatedwithincreasedtoxicity:resultsfromaEuropeanOrganisationforResearchandTreatmentofCancerGenito-UrinaryGroupPhaseIIITrial.EurUrol2003;44:429–34.19.LammDL,DeHavenJI,ShriverJ,SarosdyMF.ProspectiverandomizedcomparisonofintravesicalwithpercutaneousbacillusCalmette–GuerinversusintravesicalbacillusCalmette–Guerininsuperficialbladdercancer.JUrol1991;145(4):738–40.20.LufteneggerW,AckermannDK,FutterliebA,etal.IntravesicalversusintravesicalplusintradermalbacillusCalmette–Guerin:aprospectiverandomizedstudyinpatientswithrecurrentsuperficialbladdertumors.JUrol1996;155(2):483–7.21.RintalaE,JauhiainenK,AlfthanO,etal.Intravesicalchemotherapy(MitomycinC)versusimmunotherapy(bacillusCalmette–Guérin)insuperficialbladdercancer.EurUrol1991;20:19–25.22.LammDL,BlumensteinBA,CrawfordED,etal.RandomizedintergroupcomparisonofbacillusCalmette–GuerinimmunotherapyandmitomycinCchemotherapyprophylaxisinsuperficialtransitionalcellcarcinomaofthebladderasouthwestoncologygroupstudy.UrolOncol:SeminOriginalInvest1995;1(3):119–26.23.KregeS,GianiA,MeyerGR,OttoT,RubbenH.Arandomizedmulticentertrialofadjuvanttherapyinsuperficialbladdercancer:transurethralresectiononlyversustransurethralresectionplusmitomycinCversustransurethralresectionplusbacillusCalmette–Guerin.ParticipatingClinics.JUrol1996;156(3):962–6.24.WitjesWPJ,WitjesA,OosterhofGO,DebruyneFMJ.UpdateontheDutchCooperativeTrial:mitomycinversusbacillusCalmette–Guérin–TiceversusbacillusCalmette–GuérinRIVMinthetreatmentofpatientswithpTA–pT1papillarycarcinomaandcarcinomainsituoftheurinarybladder.SeminUrolOncol1996;1:10–6.25.WitjesJA,vanderMeijdenAP,ColletteL,etal.Long-termfollow-upofanEORTCrandomizedprospectivetrialcomparingintravesicalbacilleCalmette–Guerin–RIVMandmitomycinCinsuperficialbladdercancer.EORTCGUGroupandtheDutchSouthEastCooperativeUrologicalGroup.EuropeanOrganisationforResearchandTreatmentofCancerGenito-UrinaryTractCancerCollaborativeGroup.Urology1998;52(3):403–10.26.MalmstromPU,WijkstromH,LundholmC,WesterK,BuschC,NorlenBJ.5-YearfollowupofarandomizedprospectivestudycomparingmitomycinCandbacillusCalmette–Guerininpatientswithsuperficialbladdercarcinoma.Swedish–NorwegianBladderCancerStudyGroup.JUrol1999;161(4):1124–7.27.MangiarottiB,TrinchieriA,DelNeroA,MontanariE.Arandomizedprospectivestudyofintravesicalprophylaxisinnon-muscleinvasivebladdercanceratintermediateriskofrecurrence.Mitomycinchemotherapyvs.BCGimmunotherapy.ArchivioItalianodiUrologiaeAndrologia2008;80(4):167–71.28.OjeaA,NogueiraJL,SolsonaE,etal.Amulticentre,randomisedprospectivetrialcomparingthreeintravesicaladjuvanttherapiesforintermediate-risksuperficialbladdercancer:low-dosebacillusCalmette–Guerin(27mg)versusverylow-dosebacillusCalmette–Guerin(13.5mg)versusmitomycinC.EurUrol2007;52(5):1398–406.29.SekineH,OhyaK,KojimaS,IIgarashiK,FukuiI.EquivalentefficacyofmitomycinCplusdoxorubicininstillationtobacillusCalmette–Guerintherapyforcarcinomainsituofthebladder.IntJUrol2000;8(9):483–6.30.ShelleyMD,WiltTJ,CourtJ,ColesB,KynastonH,MasonMD.IntravesicalbacillusCalmette–GuerinissuperiortomitomycinCinreducingtumourrecurrenceinhigh-risksuperficialbladdercancer:ameta-analysisofrandomizedtrials.BJUInt2004;93(4):485–90.31.MalmstromP-U,SylvesterRJ,CrawfordDE,etal.Anindividualpatientdatameta-analysisofthelongtermoutcomeofrandomisedstudiescomparingintravesicalmitomycinCversusbacillusCalmette–Guerinfornon-muscleinvasivebladdercancer.EurUrol2009;56:247–56.32.FriedrichMG,PichlmeierU,SchwaiboldH,ConradS,HulandH.Long-termintravesicaladjuvantchemotherapyfurtherreducesrecurrenceratecomparedwithshort-termintravesicalchemotherapyandshort-termtherapywithBacillusCalmette–Guerin(BCG)inpatientswithnon-muscle-invasivebladdercarcinoma.EurUrol2007;52(4):1123–30.33.BohleA,BockPR.IntravesicalbacilleCalmette–GuerinversusmitomycinCinsuperficialbladdercancer:formalmeta-analysisofcomparativestudiesontumorprogression.Urol2004;63(4):682–6.34.DiStasiSM,GiannantoniA,GiuriolaA,etal.SequentialBCGandelectromotivemitomycinversusBCGaloneforhigh-risksuperficialbladdercancer:arandomisedcontrolledtrial.LancetOncol2006;7:43–51.35.SylvesterRJ,vanderMeijdenAP,WitjesJA,etal.BacillusCalmette–Guerinversuschemotherapyfortheintravesicaltreatmentofpatientswithcarcinomainsituofthebladder:ameta-analysisofthepublishedresultsofrandomizedclinicaltrials.JUrol2005;174(1):86–91.36.RintalaE,JauhiainenK,KaasinenE,NurmiM,AlfthanO.AlternatingmitomycinCandbacillusCalmette–Guerininstillationprophylaxisforrecurrentpapillary(stagesTatoT1)superficialbladdercancer.FinnbladderGroup.JUrol1996;156(1):56–9.37.Gelabert-masA,ToroO,SchwaiboldH,ConradS,HulandH.AprospectiveandrandomisedstudyofrecurrenceandprogressionofsuperficialbladdercancertreatedwithmitomycincaloneversusalternatingmitomycincandBCGtherapy.ArchivosEspanolesdeUrologia1993;46:397–482.38.KaasinenE,HansW,MalmstromPU,etal.Alternatingmitomycinc,BCGinstillationsversusBCGaloneinthetreatmentofcarcinomainsituoftheurinarybladder.ANordicstudy.EurUrolSuppl2002;1:101.39.WitjesJA,CarisCT,MunganN,DebruyneFM,WitjesWP.ResultsofarandomizedphaseIIItrialofsequentialintravesicaltherapywithmitomycinCandbacillusCalmette–GuerinversusmitomycinCaloneinpatientswithsuperficialbladdercancer.JUrol1998;5:1668–71.40.SolsonaE,GonzalezM,FerandezJM,UndaM,BernuyC,PertusaC.RandomtrialcomparingintravesicalchemotherapywithMMC–BCGvsintravesicalBCGinpatientswithintermediate–highrisksuperficialbladdercancer.EfficacyevaluationofCUETOtrialno93008.EurUrol2002;1(1):101.41.KaasinenE,RintalaE,PereAK,etal.WeeklymitomycinCfollowedbymonthlybacillusCalmette–Guerinoralternatingmonthlyinterferon-alpha2BandbacillusCalmette–Guerinforprophylaxisofrecurrentpapillarysuperficialbladdercarcinoma.JUrol2000;1(1):47–52.42.KhannaO,CottoneR,SonDL,etal.SuperficialbladdercancertreatedbyintravesicalbacillusCalmette–GuerinorAdriamycin:multicenterstudyinterimreport.Urology1997;15(6):520–8.43.LammDL,BlumensteinBA,CrawfordED,etal.ArandomizedtrialofintravesicaldoxorubicinandimmunotherapywithbacilleCalmette–Guérinfortransitional-cellcarcinomaofthebladder.NewEnglJMed1991;17:1205–9.44.Martinez-PineiroJ,JimenezLeonAJ,Martinez-PineiroJrL,etal.BacillusCalmette–Guerinversusdoxorubicinversusthiotepa:arandomizedprospectivestudyin202patientswithsuperficialbladdercancer.JUrol1990;143(3):502–6.45.HinotsuS,AkazaH,AsakaS,etal.SustainedprophylacticeffectofintravesicalbacilleCalmette–Guérinforsuperficialbladdercancer:asmoothedhazardanalysisinarandomizedprospectivestudy.Urology2006;3:545–9.46.ChengCW,ChanSFP,ChanLW,etal.Twelve-yearfollowupofarandomizedprospectivetrialcomparingbacillusCalmette–Guerinandepirubicinasadjuvanttherapyinsuperficialbladdercancer.IntJUrol2005;12(5):449–55.47.deReijkeTM,KurthKH,SylvesterRJ,etal.BacillusCalmette–Guerinversusepirubicinforprimary,secondaryorconcurrentcarcinomainsituofthebladder:resultsofaEuropeanOrganizationfortheResearchandTreatmentofCancer-Genito-UrinaryGroupPhaseIIITrial(30906).JUrol2005;174(3):405–9.48.Ali-El-DeinB,NabeehBA,IsmailEH,GhoneimMA.SequentialbacillusCalmette–GuerinandepirubicinversusbacillusCalmette–Guerinaloneforsuperficialbladdertumors:arandomizedprospectivestudy.JUrol1999;162(2):339–42.49.vanderMeijdenAP,BrausiM,ZambonV,etal.Intravesicalinstillationofepirubicin,bacillusCalmette–GuerinandbacillusCalmette–GuerinplusisoniazidforintermediateandhighriskTa,T1papillarycarcinomaofthebladder:aEuropeanOrganizationforResearchandTreatmentofCancerGenito-UrinaryGroupPhaseIIITrial.JUrol2001;166(2):476–81.50.BilenC,OzenH,AkiFT,AygunC,EkiciS,KendiS.ClinicalexperiencewithBCGaloneversusBCGplusepirubicin.IntJUrol2000;7:206–9.51.BrosmanS.ExperiencewithbacillusCalmette–Guerininpatientswithsuperficialbladdercancer.JUrol1992;128:27–30.52.RischmannP,ColombelM,MalavaudB,ChopinD.ProphylacticofloxacinetoimprovetoleranceofBCGintravesicalinstillations:arandomisedprospective,double-blind,placebo-controlled,multicentrestudyinpatientswithmidtohighrisksuperficialbladdertumours.EurUrol2005;4(Suppl.3):219.53.WitjesWP,KonigM,BoeminghausFP,etal.ResultsofaEuropeancomparativerandomizedstudycomparingoralbropirimineversusintravesicalBCGtreatmentinBCG-naivepatientswithcarcinomainsituoftheurinarybladder.EuropeanBropirimineStudyGroup.EurUrol1999;36(6):576–81.54.Jimenez-CruzJF,Vera-DonosoCD,WitherspoonB,etal.Intravesicalimmunoprophylaxisinrecurrentsuperficialbladdercancer(StageT1):multicentertrialcomparingbacilleCalmette–Guerinandinterferon-alpha.Urology1997;50(4):529–35.55.KalbleT,MohringK,IkingerU,RiedaschG,StaehlerG.IntravesicalpreventionofrecurrenceofsuperficialurinarybladdercancerwithBCGandKLH.Aprospectiverandomizedstudy.Urologe1991;30(2):118–21.56.BercovichE,DeriuM,ManferrariF,IrianniG.BCGvs.BCGplusrecombinantalpha-interferon2binsuperficialtumorsofthebladder.ArchivioItalianodiUrologia,Andrologia1995;67(4):257–60.57.AoT,UchidaT,MuramotoM,etal.PostoperativeintravesicalpirarubicinplusbacillusCalmette–Guerinwithorwithoutinterleukin2instillationtherapyforsuperficialbladdercancer.NishinihonJUrol1996;58(5):538–41.58.ScaldazzaV.MitomycinC30mgfor12monthsversusmitomycinC40mgfor6monthsintheprophylaxisofsuperficialbladdercarcinomaG1–2.MinervaUrolNephrol1996;48:203–6.59.TolleyDA,ParmarMK,GrigorKM,etal.TheeffectofintravesicalmitomycinConrecurrenceofnewlydiagnosedsuperficialbladdercancer:afurtherreportwith7yearsoffollowup.JUrol1996;155(4):1233–8.60.NewlingDWW,KleinjanJH.Aprospective,randomised,multicenter,phaseIIIclinicaltrialoftheeffectofdifferentinitialtherapyregimesandmaintenanceM.D.Shelleyetal./CancerTreatmentReviews36(2010)195–205205prophylaxisinsuperficialbladdercancer,usingmitomycinC.BrJUrol1997;80(Suppl.2):31.61.SchwaiboldH,PichlmeierU,KlingenbergerHJ,HulandH.Long-termfollow-upofcytostaticintravesicalinstillationinpatientswithsuperficialbladdercarcinoma.Isshort-term,intensiveinstillationbetterthanmaintenancetherapy?EurUrol1997;31(2):153–9.62.HeneyNM,KoontzWW,BartonB,etal.IntravesicalthiotepaversusmitomycinCinpatientswithTa,T1andTIStransitionalcellcarcinomaofthebladder:aphaseIIIprospectiverandomizedstudy.JUrol1988;140(6):1390–3.63.MalmstromPU.Arandomizedcomparativedose-rangingstudyofinterferonandmitomycin-Casaninternalcontrolinprimaryorrecurrentsuperficialtransitionalcellcarcinomaofthebladder.BJUInt2002;(7):681–6..TsushimaT,NasuY,AbekiN,etal.Prophylacticintravesicalinstillationtherapyinpatientswithsuperficialbladdercancer.Resultsofarandomizedprospectivestudy.JpnJUrol1992;83(8):1314–21.65.AkazaHK,KoisoK,KotakeT,etal.Long-termresultsofintravesicalchemoprophylaxisofsuperficialbladdercancer:experienceoftheJapaneseUrologicalCancerResearchGroupforAdriamycin.CancerChemotherPharmacol1992;30(Suppl.):S15–20.66.FaiolaV,MontellaL,AddeoR,etal.Intravesicalgemicitabineversusmitomycinforrecurrentsuperficialbladdertumour(stagespTAandpT1):arandomisedprospectivestudy.EurJCancer2008;Suppl.6:123–26.67.AuJLS,BadalamentRA,WientjesMG,etal.MethodstoimproveefficacyofintravesicalmitomycinC:resultsofarandomizedphaseIIItrial.JNatlCancerInst2001;93(8):597–604.68.ColomboR,DaPozzoLF,SaloniaA,etal.Multicentricstudycomparingintravesicalchemotherapyaloneandwithlocalmicrowavehyperthermiaforprophylaxisofrecurrenceofsuperficialtransitionalcellcarcinoma.JClinOncol2003;21(23):4270–6.69.DiStasiSM,GiannantoniA,StephenRL,etal.IntravesicalelectromotivemitomycinCversuspassivetransportmitomycinCforhighrisksuperficialbladdercancer:aprospectiverandomizedstudy.JUrol2003;170(3):777–82.70.OosterlinckW,KurthKH,SchroderF,BultinckJ,HammondB,SylvesterR.AprospectiveEuropeanOrganizationforResearchandTreatmentofCancerGenitourinaryGrouprandomizedtrialcomparingtransurethralresectionfollowedbyasingleintravesicalinstillationofepirubicinorwaterinsinglestageTa,T1papillarycarcinomaofthebladder.JUrol1993;149(4):749–52.71.GudjonssonS,AdellL,MerdasaF,etal.Shouldallpatientswithnon-muscle-invasivebladdercancerreceiveearlyintravesicalchemotherapyaftertransurethralresection?Theresultsofaprospectiverandomisedmulticentrestudy.EurUrol2009;55:773–80.72.BohleA,LehyH,FreiC.Singlepostoperativeinstillationofgemcitabineinpatientswithnon-muscleinvasivetransitionalcellcarcinomaofthebladder:arandomised,double-blind,placebocontrolled,phaseIIImulticentrestudy.EurUrol2009;56:495–503.73.SylvesterRJ,OosterlinckW,vanderMeijdenAP.AsingleimmediatepostoperativeinstillationofchemotherapydecreasestheriskofrecurrenceinpatientswithstageTaT1bladdercancer:ameta-analysisofpublishedresultsofrandomizedclinicaltrials.JUrol2004;171(6Pt1):2186–90.